Bug 1628107

Is there any chance there are two catalog operator pods running (anywhere in the cluster)? I have seen this issue when we accidentally had two catalog operators running at once.

I moved the severity to low since InstallPlan application is idempotent (no errors will occur from having two created for you)

Evan,

No, only one catalog operator running at that time.

> I moved the severity to low since InstallPlan application is idempotent (no errors will occur from having two created for you)

Yes, no errors. But, it's confusing for users. I think we should find the root cause and fix it although it's a little problem.

I encounter this issue again, details:
[root@qe-jiazha-311-gce-1-master-etcd-1 ~]# oc get subscription
NAME                       PACKAGE              SOURCE                CHANNEL
mongodb-enterprise-qss5n   mongodb-enterprise   certified-operators   preview
[root@qe-jiazha-311-gce-1-master-etcd-1 ~]# oc get installplan
NAME                                   CSV                      SOURCE                APPROVAL    APPROVED
install-mongodboperator.v0.3.2-722mp   mongodboperator.v0.3.2   certified-operators   Automatic   false
install-mongodboperator.v0.3.2-hfzq6   mongodboperator.v0.3.2   certified-operators   Automatic   false

Only one catalog operator running.
[root@qe-jiazha-311-gce-1-master-etcd-1 ~]# oc get pods -n operator-lifecycle-manager
NAME                                READY     STATUS    RESTARTS   AGE
catalog-operator-76c846684c-pm4n9   1/1       Running   0          1h
olm-operator-5b7f7c4556-qwfhm       1/1       Running   0          1h

[root@qe-jiazha-311-gce-1-master-etcd-1 ~]# oc get csv 
NAME                     DISPLAY   VERSION   REPLACES   PHASE
mongodboperator.v0.3.2   MongoDB   0.3.2                Succeeded
[root@qe-jiazha-311-gce-1-master-etcd-1 ~]# oc get pods
NAME                                           READY     STATUS    RESTARTS   AGE
mongodb-enterprise-operator-7b7b8b9889-67trr   1/1       Running   0          1m

What about `oc get pods --all-namespaces`? I'm concerned there might be a catalog operator in another namespace somehow.

Evan,

No, only one catalog-operator running in the cluster:
[root@share-wmenghaglb3116-mez-1 ~]# oc get pods --all-namespaces | grep catalog-operator
operator-lifecycle-manager          catalog-operator-76c846684c-r6h6m               1/1       Running     0          50m

[root@share-wmenghaglb3116-mez-1 ~]# oc get pods --all-namespaces | grep operator
jian                                etcd-operator-7b49974f5b-5j9lf                  3/3       Running     0          17m
jian3                               etcd-operator-7b49974f5b-k5k6w                  3/3       Running     0          42s
openshift-monitoring                cluster-monitoring-operator-75c6b544dd-npkxx    1/1       Running     0          1d
openshift-monitoring                prometheus-operator-564dd668b-8t6cf             1/1       Running     0          1d
operator-lifecycle-manager          catalog-operator-76c846684c-r6h6m               1/1       Running     0          47m
operator-lifecycle-manager          olm-operator-5b7f7c4556-wg89n                   1/1       Running     0          47m

[root@share-wmenghaglb3116-mez-1 ~]# oc get subscription -n jian3
NAME         PACKAGE   SOURCE         CHANNEL
etcd-v58f8   etcd      rh-operators   alpha

Two installplan created by the same subscription:
[root@share-wmenghaglb3116-mez-1 ~]# oc get installplan -n jian3
NAME                                CSV                   SOURCE         APPROVAL    APPROVED
install-etcdoperator.v0.9.2-7w9mg   etcdoperator.v0.9.2   rh-operators   Automatic   false
install-etcdoperator.v0.9.2-gq2b9   etcdoperator.v0.9.2   rh-operators   Automatic   false
[root@share-wmenghaglb3116-mez-1 ~]# oc get csv -n jian3
NAME                  DISPLAY   VERSION   REPLACES              PHASE
etcdoperator.v0.9.2   etcd      0.9.2     etcdoperator.v0.9.0   Succeeded
[root@share-wmenghaglb3116-mez-1 ~]# oc get pods -n jian3
NAME                             READY     STATUS    RESTARTS   AGE
etcd-operator-7b49974f5b-k5k6w   3/3       Running   0          2m


[root@share-wmenghaglb3116-mez-1 ~]# oc get pods
NAME                                READY     STATUS    RESTARTS   AGE
catalog-operator-76c846684c-r6h6m   1/1       Running   0          51m
olm-operator-5b7f7c4556-wg89n       1/1       Running   0          51m
[root@share-wmenghaglb3116-mez-1 ~]# oc rsh olm-operator-5b7f7c4556-wg89n
sh-4.2$ olm -version
OLM version: 0.6.0
git commit: 3df6bea

image: registry.reg-aws.openshift.com:443/openshift3/ose-operator-lifecycle-manager:v3.11
      imageID: docker-pullable://registry.reg-aws.openshift.com:443/openshift3/ose-operator-lifecycle-manager@sha256:a706b51ae87dc9faa98b484ab06bf8ad2f4c3283de0c024b253f273c981318cc

It seems like a bug then, but I've moved it to 4.0 since it doesn't affect users at all.

*** Bug 1633885 has been marked as a duplicate of this bug. ***

I don't encounter this issue anymore for a long by using the OLM in OCP 4.0.
So, verify it. Please feel free to reopen it if you encounter this issue.
Current version:
mac:aws-ocp jianzhang$ oc exec  olm-operator-5fdc6d559f-2x7zp -- olm -version
OLM version: 0.8.0
git commit: c53c51a

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2019:0758